Care2Report: Automated Medical Reporting

During my bachelor thesis project I worked in a team that developed a prototype for a research program called Care2Report at Utrecht University. The project culminated in a contribution to a paper, that has since been presented at the Hawaii International Conference on System Sciences.

The Care2Report program aims to develop a platform that facilitates automated medical reporting. A conceptual architecture was developed prior to our project, it breaks down the reporting into multiple distinct steps. Specifically it includes a specific ‘semantic interpretation’ step, where data (natural language and/or video from a medical consultation) is converted into concrete facts in an OWL ontology, before the report is generated.

During the project I worked mostly on the report generation part, extending the open-source NaturalOWL software for this purpose. Since then I have become increasingly interested in the semantic interpretation step. Specifically what a neural model designed to perform this task might look like. I intend to write a blog post about my thoughts on this in future.